Horizontal Connections for Precast
Concrete Shear Wall Panels Under
Cyclic Shear Loading
Khaled A. Soudkl
Ph.D., P.Eng.
Adjunct Assistant Professor and
NSERC Postdoctoral Fellow
Department of Civil Engineering
Queen's University
Kingston, Ontario, Canada
Jeffrey S. West
Ph.D. Candidate
Department of Civil Engineering
University of Texas at Austin
Austin, Texas
Sami H. Rizkalla
Ph.D., P.Eng.
Professor
Department of Civil and
Geological Engineering
University of Manitoba
Winnipeg, Manitoba, Canada
Bruce Blackett, P.Eng.
Vice President and
General Manager
Con-Force Structures Ltd.
Winnipeg. Manitoba, Canada
The performance of precast concrete
loadbearing shear wall panel structures
during an earthquake relies on the behavior
and integrity of the connections between
the panels. Design of these structures
requires the ability to predict the behavior of
the connections. This paper presents
results of an experimental study of
horizontal connections for precast wall
panels subjected to reversed cyclic shear
deformations combined with simulated
gravity loads normal to the connection. The
study considered typical connection
configurations currently used in practice. The
influence of mild steel reinforcement, post-
tensioning and shear keys was investigated.
Experimental results were used to determine
the cyclic behavior of the connections and to
identify the contribution of the connection
components. Simple rational models were
proposed to predict the limit states of the
connections. Two numerical design
examples are included to illustrate the
proposed models. Based on the study,
design recommendations are presented.
he precast concrete loadbearing shear wall panel sys-
T tem is an economical construction system for low,
medium and high rise structures. This form of multi-
story structural system is composed of large precast con-
crete wall panels and floor panels arranged to form box-like
apartment or office size spaces. The walls of the structure
carry both the gravity and lateral loads on the structure and
an intermediate frame is not required. The wall panels are
typically one story in height and are connected both hori-
PCIJOURNAL
zontally and vertically to adjacent pan-
els at each story.
Vertical connection between panels
may be provided by welded or bolted
shear connectors. At the horizontal
connection, both shear and vertical
forces must be transferred. Vertical
continuity is normally achieved using
post-tensioning or with welded or me-
chanically spliced mild steel continu-
ity bars. Mechanical shear connectors
or multiple shear keys are often used
to enhance shear transfer at the con-
nection. During erection, a gap is left
between adjacent panels for alignment
purposes. In North American practice,
this space is fllied with drypack grout
to complete the connection.
The advantage of this type of struc-
tural system lies in its high quality pre-
cast elements and rapid cost-effective
erection. Cost savings compared to
other forms of construction can be
achieved through the use of large wall
panels with minimal connections.
However, this approach places large
demands on the connections between
the elements. In general, the connec-
tions have lower strength and stiffness
than the wall panels and form planes of
weakness. Because the horizontal con-
nections of the structure must transmit
both the gravity and lateral loads on
the structure, they playa critical role in
sustaining structural stability and in-
tegrity. Consequently, it is important to
fully understand their behavior.
In seismically active regions, the
use of precast concrete structural sys-
tems has primarily been limited to low
rise structures. The main reason for
their exclusion from use in medium
and high rise applications is the lack
of knowledge of how this type of con-
struction will perform under seismic
loading conditions. Coupled with this
shortage of technical information, the
current building codes and design
guidelines of North America l •7 do not
specifically address the use of precast
concrete in seismic zones. Rather, pre-
cast concrete is required to comply
with code provisions developed for
cast-in-place concrete that do not con-
sider the unique behavior of precast
concrete and in most cases eliminate
the competitive advantages of its use.
Acceptance and competitiveness of
precast concrete loadbearing shear
May-June 1996
wall panel systems in the seismic
zones of North America depend on
Wlderstanding the behavior of the con-
nections and development of guide-
lines and recommendations for the de-
sign and use of these types of
connections.
To achieve this goal, a multi-phase
research program was undertaken to
investigate the behavior of typical hor-
izontal connection configurations
under reversed cyclic shear loading
equivalent to earthquake loading con-
ditions. Connections were also sub-
jected to axial stresses normal to the
connection to simulate gravity loading
conditions. A companion research pro-
grama.lo was conducted in parallel to
consider the cyclic flexure and shear
behavior of currently used connections
and innovative connection details be-
lieved to enhance the ductility and/or
capacity for energy dissipation. The
intent of the research is to fully defme
the horizontal connection behavior to
promote development of precast shear
wall panel structural systems for use
in both moderate (Zone II) and severe
seismic zones (Zones III and IV).
The scope of these research pro-
grams expands a previous five-year re-
search investigation of the monotonic
shear behavior of typical precast con-
crete shear wall connections per-
formed at the University of Manitoba
and published in the PCI JOURNAL
(Foerster et aI.,l1 Hutchinson et al.,tl
and Serrette et aI. 13).
RESEARCH
SIGNIFICANCE
The experimental program l4 de-
scribes the cyclic shear behavior of
horizontal connections incorporating
prototype connection details currently
used by the precast concrete industry.
The observed behavior was used to
identify the various limit states and
modes of failure Wlder cyclic loading.
The experimental results were used to
propose mechanisms of shear transfer
for the different connection configura-
tions and to develop simple analytical
models to predict the connection be-
havior at the various limit states. The
research study was used to introduce
shear design recommendations for
horizontal connections for precast
concrete loadbearing shear wall panel
systems in seismic zones.
EXPERIMENTAL
PROGRAM
The experimental program was de-
signed to evaluate the connection be-
havior under the effects of cyclic shear
conditions utilizing the static (mono-
tonic) test results conducted on the
same connections in the previous
study at the University of Manitoba. II .13
A total of six specimens were consid-
ered with five different connection
configurations. All connection config-
urations were tested under cyclic load-
ing. A sixth specimen was tested
under monotonic shear loading as one
connection configuration was not con-
sidered in the previous research men-
tioned above. Table 1 summarizes the
experimental program.
COnnection COnfigurations
The five connection configurations
tested are shown in Fig. 1. All connec-
tions incorporated a 20 mm e/4 in.)
thick gap between the two panels
filled with drypack grout. All vertical
continuity reinforcement or post-
tensioning was centered on the con-
nection and spaced at 900 mm (36 in.)
apart. The five connection configura-
tions are as follows:
DP: Drypack Plain Surface -
Connection DP had a plain surface re-
gion filled with drypack, as shown in
Fig. l(a). This specimen was used to
identify the contribution of the dry-
pack to the connection behavior and is
the control specimen for the study.
Typically, some form of vertical conti-
nuity element is required across the
connection.
SK: Multiple Shear Keys - The
connection interface surface for Con-
nection SK consists of five shear keys,
as shown in Fig. l(b). The length of
the shear key is 100 mm (4 in.), the
depth is 35 mm W/8 in.) and the sides
of the key are inclined at 23 degrees
from the vertical. The shear keys and
the gap between the panels are com-
pletely filled with drypack grout. No
vertical reinforcement is provided
across the connection interface in
order to study the influence of the
2
shear keys alone.
RW: Reinrorcing Bars Welded to
Steel Angle - This connection is
commonly used by the Canadian pre-
cast industry. The connection consists
of a plain surface region with two
25M (#8), Grade 400W (60 ksi) mild
steel bars. The straight bars protruding
from the upper panel are welded to a
75 x 75 x 10 mm (3 x 3 x liz in.),
Grade 300 (40 ksi) steel angle in an
exposed pocket in the lower panel, as
shown in Fig. l(c). After welding, the
gap between the panels is filled with
drypack grout.
PTS: Post·Tensioned Strand -
Vertical continuity in Connection PTS
is provided using two 12.7 mm (1/2 in.)
diameter, Grade 1860 (270 ksi) seven-
wire strands. The gap between the
panels was filled with drypack grout
and allowed to cure for seven days
prior to post-tensioning. The strands
were tensioned to 60 percent of their
ultimate strength to produce an aver-
age stress of 1.2 MPa (174 psi) normal
to the connection. The strands were
placed inside galvanized steel ducts
cast in the panels. After post-tension-
ing, the ducts were filled with an ex-
pansive grout The PTS configuration
is shown in Fig. l(d).
PTB: Post-Tensioned Bar - Con-
nection PTB was post-tensioned using
two 15.8 mm (S/8 in.) diameter, Grade
1080 (157 ksi) threaded prestressing
bars. Similar to Connection PTS, the
bars were prestressed seven days after
drypacking to produce an average
compressive stress of 1.2 MPa (174
psi) normal to the connection. The
bars were placed inside galvanized
steel ducts. After post-tensioning, the
ducts were filled with expansive grout.
The PTB configuration is shown in
Fig.1(d).
Test Specimen and Setup
The overall geometry of the test
specimens was selected to induce
shear loads at the interface of the con-
nection as shown in Fig. 2. The speci-
men consisted of an upper panel and a
lower panel joined at a 1200 mm long
and 150 mm wide connection (48 in.
long x 6 in. wide). The upper panel
had an inverted U-shape with double
corbels to allow the application of the
3 PCI JOURNAL
reversed cyclic shear loading.
In addition to shear loading, all
specimens were subjected to a uniform
vertical stress, perpendicular to the
connection, to simulate the effects of
gravity loads for a typical ten-story
structure of this type. The magnitude
of this stress was 2 MPa (290 psi) for
all specimens except the drypack grout
only connection (Connection DP),
which was subjected to a vertical
stress of 4 MPa (580 psi).
The specimens were tested using a
structural steel loading frame as
shown in Fig. 3. The lower wall panel
of the specimen was fixed to the floor
of the structures laboratory by post-
tensioning against two reaction abut-
ments, as shown in Fig. 3. The cyclic
shear loading was applied to the upper
concrete panel using a push/pull load-
ing yoke and a 1000 leN (225 kips) ca-
pacity MrS closed loop cyclic actua-
tor. The configuration of the upper
panel and the loading yoke allowed
the application of a concentric shear
load to the connection region.
The equivalent gravity load, normal
to the connection, was applied using
an independent system designed to
allow free displacement in the direc-
tion of the applied shear load. The sys-
tem utilized two hydraulic jacks,
which maintained a constant vertical
load normal to the connection using a
regulator valve and a hydraulic pump.
The specimen was instrumented
with linear variable differential trans-
ducers (L VDTs) and mechanical dial
gauges to measure the relative hori-
zontal slip between the two panels and
the vertical deformation of the dry-
pack grout Strain gauges were placed
on the post-tensioning bars or mild
steel continuity reinforcement at the
connection level.
Test Procedure
Prior to application of the simulated
gravity load normal to the connection,
initial readings were recorded for all
instrumentation. The load normal to
the connection was then applied and
maintained for the duration of testing.
The test proceeded by applying the
shear loading following a quasi-static,
cyclic loading history. A combination
of load and displacement control was
used to determine the behavior of the
connection in the elastic and inelastic
ranges. Each load/displacement level
was maintained for three fully re-
versed cycles.
Initially, loading was applied under
load control with increments of 50 or
100 leN (11.2 or 22.4 kips). Following
the initiation of slip at the connection,
the behavior of the connection became
inelastic and loading was switched to
displacement control. In this program,
the initiation of slip constituted a rela-
tive slip of 0.1 mm (0.004 in.) be-
tween the two panels. Displacement
increments were taken as 1.0 mm
(0.04 in.) of relative horizontal slip
measured between the panels at the
connection.
The test was continued under dis-
placement control until failure of the
connection occurred, defined as a 20
percent reduction in the shear resis-
tance of the connection. After failure,
the test was concluded by applying
several single cycles at increasing slip
magnitudes to determine the residual
strength of the connection. The gen-
eral form of the loading history is
shown in Fig. 4.
EXPERIMENTAL RESULTS
Connection Behavior
The general behavior of the connec-
tions tested under cyclic shear loading
was quite similar for the different
connection configurations considered
in this investigation with the excep-
tion of the shear key configuration,
Connection SK. The shear resistance-
slip behaviors under reversed cyclic
shear loading for the five connection
configurations are shown in Figs. 5(a)
to 5(e).
For all tested connections, the initial
response was extremely stiff and elas-
tic up to the initiation of the first slip
at the interface of the drypack grout
and the concrete panel. The initiation
of slip occurred gradually for all con-
nection configurations with the excep-
tion of the connection with drypack
grout only, Connection DP.
Following the initiation of slip, the
behavior of the drypack grout connec-
tion, Connection DP, and the post-
tensioned strand connection, Connee-
 tion PTS, was consistent, resulting in
almost rectangular shear resistance-
slip hysteresis loops, as sho~ in Fi~s.
Sea) and Sed). The connectIon W.Ith
mild steel continuity bars, Connection
RW, exhibited parallelogram-shaped
hysteresis loops with increasing shear
resistance at larger slip magnitudes, as
shown in Fig. S(c). The behavior of
the post-tensioned bar connection,
Connection PTB, produced parallelo-
gram-shaped hysteresis loops with de-
creasing shear resistance as the ap-
plied slip magnitude was increased, as
shown in Fig. 5(e).
The behavior of the shear key con-
nection Connection SK, was somewhat
differe~t, as is evident in Fig. 5(b). Ini-
tiation of slip occurred at a load level of
400 leN (50 kips). The connection re-
sisted a maximum applied load of 850
leN ( 191 kips) with a limited slip of less
than 1 mm (0.04 in.). This maximum
capacity was considerably higher than
the maximum measured capacity for all
of the tested plain surface connection
configurations.
Failure of Connection SK occurred
suddenly due to simultaneous cracking
through the drypack grout in ea~h of
the five shear keys. This resulted 10 an
immediate loss of shear resistance as
shown in Fig. 5(b). Following failure
of the drypack grout within the shear
keys, a slip interface developed along
the length of the connection and the
behavior stabilized, producing almost
rectangular shear resistance-slip hys-
teresis loops similar to the plain sur-
face connections.
The behavior of the post-tensioned
bar connection tested under static
loading, Connection PTB-S, is shown
in Fig. S(t). Slip of the connection ini-
tiated at approximately the same load
level as Connection PTB, tested under
cyclic loading conditions. Howev~r,
under static loading, the shear reSIS-
tance steadily increased as the applied
slip magnitude increased. . bars at the connection level at the end
The trend of increasing shear reSIS-
tance of Connection PTB-S continued
up to a very significant slip in excess
of 35 mm (1.4 in.). At this stage, there
was no measurable reduction in the
thickness of the drypack grout. This
behavior is consistent with behavior
reported in the previous studies under
monotonic loading ll -11 and is quite dif-
May-June 1996
ferent from the behavior of the same
connection configuration under re-
versed cyclic load (Connection PTB).
Failure Modes
The mode of failure for all of the
connections tested under cyclic load-
ing was due to sudden crushing and
spalling of the drypack grout ~co~­
panied by a significant reduction 10
the thickness of the grout. The correla-
tion between crushing of the drypack
grout and sudden loss of shear resis-
tance is shown in Fig. 6. The results
shown in this figure are from the con-
nection with post-tensioned strands,
Connection PTS, and are representa-
tive of the behavior observed for other
connections. Prior to any notable loss
of shear resistance, the reduction in
grout thickness was insignific~t [less
than 1 mm (0.04 in.)]. After faIlure of
the drypack grout, the hysteresis loops
stabilized at a reduced shear resistance
and the drypack grout was virtually
ground into powder.
The effect of this occurrence on the
behavior of the connection depends on
the resistance mechanisms involved.
While Connection DP experienced
only a minor loss of shear resistance,
the other configurations displayed a
strength reduction in the order of
50 percent, as shown in Figs. 5(a)
through (e).
For Connection DP, crushing of the
drypack grout represents a reduction
in the coefficient of friction. Because
this is the only resistance mechanism
for this configuration, the reduction in
shear resistance was not substantial.
For the connection with mild steel
continuity bars, Connection RW,
crushing of the drypack caused trans-
fer of the applied load normal to the
connection to the two continuity bars.
This behavior was confirmed by the
out-of-plane buckling of one of the
of the test. Because the vertical load
was no longer acting entirely on the
interface surface of the connection,
shear resistance provided by friction
was also reduced. The connection con-
dition after crushing of drypack is
shown in Fig. 7(a).
The post-tensioned connections,
PTB and PTS, displayed a consider-
able loss of shear resistance upon
crushing of the grout. As in the case of
Connection DP, the coefficient of fric-
tion is reduced. In addition, the crush-
ing of the grout and subsequent reduc-
tion in the grout thickness could
induce a substantial loss of the pre-
stressing force Therefore, both the co-
efficient of friction and the equivalent
normal force on the connection due to
post-tensioning are reduced, resul~g
in a substantial loss of shear reSIS-
tance. The condition of the connection
after crushing of drypack for Connec-
tion PTB is shown in Fig. 7(b).
For Connection SK, the shear resis-
tance mechanism before failure of the
drypack is largely provided by direct
bearing within the shear keys. The .de-
struction of this resistance mechanIsm
occurred suddenly for Connection SK
due to simultaneous cracking of the
drypack in all the ~hear key~, as
shown in Fig. 7(c). ThIS resulted 10 an
immediate drop in the shear resis-
tance as illustrated in Fig. 5(b). Fol-
10wU;g failure of the drypack within
the shear keys, a slip interface devel-
oped along the connection le~gth as
shown in Fig. 7(d). The behaVIor sta-
bilized with almost rectangular shear-
slip hysteresis loops.
DISCUSSION OF
TEST RESULTS
The test results are compared using
the envelope of the shear resistance-
slip cyclic response. The results ~f
Connection DP are scaled to allow di-
rect comparison with all other speci-
mens tested at a stress of 2 MPa (290
psi) normal to the connection.
General Behavior
The test -results suggest that the
cyclic shear behavior of the conn~c­
tions could be described in three dIS-
tinct stages. The ftrst stage descri~s
the perfectly elastic, stiff behaVIOr
prior to initiation of slip between the
drypack grout and the panels. . .
The second stage describes melasttc
behavior after the initiation of slip.
During this stage, the drypack grout
remains intact and some increase or
decrease in shear resistance may occur
as the applied slip magnitude is in-
4
creased. The shear key connection,
Connection SK, exhibited a dramatic
increase in shear resistance with lim-
ited slip during this stage due to the
presence of the shear keys.
The third stage of behavior is initi-
ated by sudden and extensive crushing
and spalling of the drypack grout, ac-
companied by a significant reduction
in shear resistance. This generalized
behavior is illustrated in Fig. 8.
Effect of Cyclic load
Deterioration of the drypack grout
due to cyclic loading introduces an ad-
ditional limit state beyond that ob-
served under monotonic loading con-
ditions. Initially, the grinding action of
the cyclic loading caused a gradual re-
duction in the drypack grout thickness,
producing increased compression
forces in the continuity bars (Connec-
tion RW) and a loss of prestress in the
specimen with post-tensioned bars
(Connection PTB). At failure, cyclic
loading caused significant crushing
and spalling of the drypack grout. This
behavior, described previously as
Stage III of cyclic behavior, did not
occur under monotonic loading. II - 14
The cyclic loading effect is evident by
comparing companion prestressed bar
connections tested under static and
cyclic loading conditions, as shown in
Fig. 9.
Effect of Mild Steel
Reinforcement
The presence of mild steel continu-
ity reinforcement enhanced the shear
resistance relative to the connection
with drypack grout only (Connection
DP). The resistance of the mild steel
reinforcement to deformation provides
an additional component of shear re-
sistance due to the dowel action of the
reinforcing bars. This mechanism is
directly related to the magnitude of the
slip. As the slip is increased, the local
deformation of the bars and, therefore,
the resistance of the bars became more
significant, increasing the shear resis-
tance of the connection. The increased
shear resistance provided by the mild
steel continuity reinforcement is evi-
dent in Fig. 10 where envelopes of
cyclic response for Connections DP
and RW are compared.
5 PCI JOURNAL
Effect of Prestressed tance in comparison to the interface
Reinforcement for Connection OP, consisting of
smooth concrete-to-drypack regions
Prestressing of the connection using
only.
bars or strands enhances the frictional
resistance by increasing the effective
stress normal to the connection. As a SHEAR RESISTANCE
result, the overall shear resistance is
MECHANISMS
increased, as is evident by comparing
On the basis of the observed experi-
the shear resistance vs. slip envelopes
mental behavior and the measured re-
for the prestressed connections (Con-
sults, mechanisms of shear transfer
nections PTB and PTS) and the dry-
were proposed for each connection
pack grout only connection (Connec-
configuration. Where possible, mathe-
tion DP) in Fig. 11. At failure,
matical models were proposed for the
significant crushing of the drypack
prediction of the connection capacity
grout could lead to a total loss of pre-
at the various limit states of cyclic be-
stressing. Consequently, the additional
havior. However, it should be men-
frictional resistance due to post-
tioned that no attempt was made in the
tensioning will be eliminated, result-
models to evaluate the slip magnitude
ing in a shear resistance similar to
for the corresponding limit state. The
Connection DP, as shown in Fig. 11.
models presented predict the shear
strength of the connection at various
Effect of Shear Keys limit states.
The presence of multiple shear keys
significantly enhances the shear-slip Connection With Mild Steel
resistance. This is illustrated by com- Reinforcement
paring the response of the shear key
The shear resistance mechanism for
connection, Connections SK and DP
in Fig. 12. The maximum shear resis~
plain surface connections with mild
steel continuity reinforcement is pro-
tance of Connection SK was more
vided by two components: (1) friction
than three times that of Connection
at the drypack grout-to-panel interface
OP. The increase in the shear resis-
and (2) resistance of the continuity
tance of Connection SK is mainly due
bars to deformation. The initial fric-
to the transfer of shear by direct bear-
tional resistance, Vfh is provided by
ing of the drypack grout within the
the net gravity stress acting on the
shear keys.
connection, O'nl' as shown in Fig. 13.
Prior to cracking of the drypack
The remainder of the gravity load is
grout within the shear keys, the slip
resisted by the continuity reinforce-
was limited to less than 1 mm (0.04
ment, Fh•
in.). After cracking of the drypack
The shear resistance provided by the
grout within the shear keys, slip oc-
continuity bars does not become active
curred along the length of the connec-
until slip has occurred. Contribution of
tion and the shear resistance was pro-
the bars to the shear resistance was in
vided mainly by friction.
a progression of a flexural and kinking
The ultimate shear resistance of
mechanism. The shear resistance of
Connection SK was 40 percent higher
the bar due to the flexural mechanism
than Connection OP. The higher shear
Vhlt shown in Fig. 14(a), is dominan~
resistance of Connection SK could be
until the formation of plastic hinges in
attributed to the different slip inter-
the bars. The shear resistance of the
faces for the two connection types at
bars after the formation of plastic
ultimate_ The slip interface for Con-
hinges is due to a kinking mechanism,
nection SK is enhanced due to con-
as shown in Fig. 14(b).
finement of the drypack grout within
As horizontal slip is increased, a
the shear keys. It is assumed that this
tensile force is induced in the bars
interface, consisting of a combination
along the kinked length_ The vertical
of confined drypack-to-drypack re-
component of this force exerts a verti-
gions within the shear keys and con-
cal stress on the drypack grout and
crete-to-drypack regions elsewhere
will provide a higher frictional resis~
produces additional frictional resis-
t.ance Vj'2. This is shown in Fig. 14(b)
and is referred to as clamping action.
In addition, the horizontal component
of the tensile force in the bar, V"l' will
directly resist the applied shear load.
The proposed shear resistance
mechanisms are primarily a function
of the force within the continuity bars.
Prediction of the response based on
such a mechanism requires the predic-
tion of the force in the continuity bar
at different slip magnitudes. Because
the variation of the compressive force
in the bar cannot be rationally pre-
dicted, the approach becomes very
complicated and inconvenient for de-
sign purposes. Detailed information is
given in Ref. 14.
To avoid the necessity for prediction
of the bar forces, the following simpli-
fication may be applied. Earlier, it was
stipulated that the grinding action of
the reversed cyclic loading causes a
reduction in the thickness of the con-
nection and consequently induces
compression strains in the continuity
bars. Due to the induced precompres-
sion, the bar will not yield in tension
under cyclic load as was observed
under monotonic loading conditions.ll
Experimentally measured strains ob-
tained for Connection RW under
cyclic loading indicated that the maxi-
mum tension strain increment in the
bars due to the kinking mechanism
was limited to 40 percent of the yield
level.
Thus, for design purposes, the in-
creased frictional resistance due to
clamping action, Vf2, may be based on
the assumption of a maximum tensile
stress increase in the bars equal to 40
percent of the yield strength. It should
be noted that this value may be depen-
dent on the percentage of steel across
the connection region. Although the
horizontal component, Vh2 , can also be
calculated in this manner, examination
of the experimental results suggests
that the contribution of this mecha-
nism was negligible.
For design purposes, prediction of
the initiation of slip and the maximum
and ultimate shear resistance of the
connection may be achieved using the
general form given as follows:
Vr=V,,+Vb (I)
where V,. is the frictional resistance
May-June 1996
provided by load normal to the con- Connection With Pres1ressed
nection given by: Reinforcement
VII =J.L (1I1Ac (2) The shear resistance mechanism of
plain surface connections with post-
and V" is the frictional resistance pr0- tensioning is mainly provided by fric-
vided by the clamping action of the tion at the drypack grout-to-panel in-
bar: terface. The frictional resistance of the
Vb =J.LAJs (3) connection is due in combination to
the applied vertical gravity loads and
Therefore:
post-tensioning stresses normal to the
~ =J.L«(1,.Ac + AJs) (4) connection. The total shear resistance
could be separated into two compo-
where
nents to illustrate the contribution of
J.L =coefficient of friction the applied load normal to the connec-
=0.80 initially to 0.6 at ultimate tion and the prestressing force:
(based on test results)
(111 =compressive stress due to nor- ~= v,. + ~ (5)
mal loads
where V,. is the frictional resistance
Ac =cross-sectional area of connec-
provided by the load normal to con-
tion
nection given by:
=
As total cross-sectional area of con-
tinuity bars V,. =/.l (1,.Ac (6)
h =tensile stress increase in conti- and ~ is the frictional resistance pro-
nuity bar « 1,)
vided by post-tensioning:
Prior to initiation of slip, the compo-
nent of shear resistance due to clamp- ~ =/.l (1p A c (7)
ing action of the bar is not activated
because it is dependent on the extent of Therefore:
slip. At this stage, the shear resistance ~ =J1. «(1,. + CTp ) Ac (8)
is provided only by interface friction
where
due to gravity loads and may be pre-
J1. =coefficient of friction
dicted using Eq. (2) with J.L = 0.8.
The maximum shear resistance of
=0.80 initially to 0.6 at ultimate
(based on test results)
the connection is developed after con-
G,. = compressive stress due to nor-
siderable slip along the connection.
mal loads
Shear resistance is provided by fric-
(1p =compressive stress due to post-
tion due to gravity loads and clamping
tensioning
action resulting from deformation of
Ac =cross-sectional area of connec-
the bars. As described previously, the
tion
maximum tensile stress increase in the
For both types of post-tensioned
continuity bars due to kinking action,
connections investigated in this exper-
Is, may be taken as 40 percent of the
imental program, the shear resistance
yield strength of the bar. At this stage,
at the initiation of slip and the maxi-
the drypack grout remains intact and
mum shear resistance were very simi-
the shear resistance may be predicted
lar. At both of these limit states, the
using Eq. (4) with J1. = 0.8.
drypack grout is still intact and the
At ultimate, extensive crushing of
loss of prestressing is insignificant.
the drypack grout occurs and the
Thus, the initiation of slip and the
clamping action of the continuity rein-
maximum shear resistance of the con-
forcement is no longer reliable due to
nection may be predicted using Eq. (8)
possible buckling of the bars at the
with J1. =0.8.
connection level. In addition, the coef-
At ultimate, extensive crushing and
ficient of friction is reduced due to
spalling of the drypack grout was ob-
crushing of the drypack grout At this
served for both connection types with
stage, the shear resistance of the con-
post-tensioning. At this limit state,
nection is provided only by friction
crushing of the drypack grout was ac-
due to gravity loads and may be pre-
companied by a significant reduction
dicted using Eq. (2) with J1. =0.6.
in the thickness of the connection,
6
leading to a complete loss of prestress-
ing. As a result, the frictional resis-
tance provided by post-tensioning is
reduced to zero. Crushing of the dry-
pack grout also lowered the coefficient
of friction of the connection. Thus, the
ultimate shear resistance of the con-
nection may be predicted using Eq. (6)
with J.l =0.6.
Connection With Muhlple
Shear Keys
The observed cyclic behavior of the
connection with drypacked multiple
shear keys was not significantly differ-
ent from the behavior observed previ-
ously under monotonic loading condi-
tions. II •l l The mathematical models
proposed by Serrette et al. ll for mono-
tonic loading conditions appear to also
be applicable under cyclic loading.
The reader is referred to Ref. 13 for a
complete description of these models.
DESIGN EXAMPLES
Consider a typical connection for a
precast concrete loadbearing interior
shear wall panel at the base of a 10-
story building. The connection is 1200
mm (48 in.) long and 150 mm (6 in.)
wide with a 20 mm ('/4 in.) drypacked
thickness between the two precast
panels. The gravity load on the con-
nection from the upper stories is
equivalent to a stress, C1", of 2 MPa
(290 psi). Vertical continuity between
the two panels is provided by two con-
tinuity reinforcement at 900 mm
(36 in.) on center.
Two examples are included to show
the prediction of the shear strength for
connections with mild steel and pre-
sn:essed reinforcement. The continuity
remforcement is: (1) 25M (#8), Grade
400 MPa (60 ksi) mild steel reinforc-
ing bars and (2) 12.7 mm e/2 in.) di-
ameter seven-wire prestressing strand,
Grade 1860 (270 ksi). The characteris-
tic properties of the different materials
used in the connection are:
Concrete:
Cylinder compressive strength,
f;= 48 MPa (6.96 ksi)
Drypack grout:
Cube compressive strength,
f;= 55 MPa (7.98 ksi)
Cylinder compressive strength,
7 PCI JOURNAL
f; =0.851;= 46 MPa (6.67 ksi)
Mild Steel Bars:
Area of one bar,
=
A, 500 mm2 (0.79 sq in.)
Yield strength,
=
!, 400 MPa (60 ksi)
Ultimate strength,
=
I,. 680 MPa (98.6 ksi)
Prestressing Strand:
Area of one strand,
Api =99 mm2 (0.153 sq in.)
Yield strength,
Ipy = 0.90/p" = 1675 MPa (243 ksi)
Ultimate strength,
=
/p,. 1860 MPa (270 ksi)
Example 1: Mild Steel
Reinforcement Connection
Initially, the component of shear re-
sistance due to the bars is zero because
the slip is zero. The shear resistance at
the initiation of slip is provided by
friction and is calculated using Eq. (2):
v,. =J.l C1" Ac
=0.8(2)(180,000)
=288 kN (64.8 kips)
As discussed previously, at maxi-
mum shear resistance, the tensile
stress increase in the bars is limited to
40 percent of the yield strength.
Therefore, the maximum shear resis-
tance can be determined as follows:
v,. =J.l (C1"Ac + Ash)
=0.80[(2)(180,000)
+ (2)(500)(0.4 x 400)]
=416 kN (93.5 kips)
At failure, the severe drypack grout
deterioration causes a reduction in the
coefficient of friction from 0.8 to 0.6.
The bars buckle out-of-plane and their
contribution to the shear resistance re-
duces to zero at this stage. The shear
resistance becomes:
V,. =J.l(C1"Ac+AJs)
=0.60 [(2)(180,000) + (2)(500)(0)]
=216 kN (48.6 kips)
The predicted maximum and mini-
mum strength levels are compared to
the envelope of the measured cyclic
shear-slip behavior of mild steel con-
nection, Connection RW, with identi-
cal configuration to the example in
Fig. 15. The measured resistance at
initiation of slip for Connection RW
was within 1 percent of the predicted
value. The maximum resistance was 4
percent lower than the predicted value
and the measured ultimate strength
was 14 percent lower than the pre-
dicted value.
Example 2: Prestressed
Connection
The initiation of slip and the maxi-
mum shear resistance of a connection
post-tensioned using strands can be
predicted using Eq. (8) with a nomi-
nal coefficient of friction of J1. =0.80.
The shear resistance is determined as
follows:
v,. = J.l(C1" + C1p )Ac
= 0.80(2 + 1.2)(180,000)
= 460 kN (103.5 kips)
At failure, the severe drypack grout
deterioration caused a reduction in the
connection thickness and a complete
loss of prestressing. As a result, the
contribution from component v;,
is re-
duced to zero and the shear resistance
is provided only by friction due to the
load normal to connection, V"' with a
reduced coefficient of friction of J1. =
0.6. Thus, the shear resistance be-
comes:
V,. = J1. (C1" + C1p ) Ac
= 0.60(2 + 0) (180,000)
= 216 kN (48.6 kips)
The predicted maximum and mini-
mum strength levels are compared to
the envelope of the measured cyclic
shear-slip behavior for the connection
post-tensioned using strands, Connec-
tion PTS in Fig. 16. This connection
has an identical configuration to the
example. The measured maximum
strength was 15 percent higher than
the predicted value and the measured
ultimate strength was 4 percent higher
than the predicted value.
CONCLUSIONS
Based on the results of this study,
the following conclusions can be
drawn:
1. The cyclic behavior of the con-
nection could be identified by three
distinct limit states: (a) elastic and
very stiff behavior prior to initiation
of slip; (b) inelastic behavior with sta-
-------------
ble hysteresis before significant joint
deterioration; and (c) failure with sig-
nificant joint deterioration and more
than 20 percent reduction in the shear
resistance.
2. The mode of failure for all con-
nection configurations tested under
cyclic loading was due to significant
crushing and spalling of drypack
grout. This introduced an additional
limit state beyond that observed for
identical specimens tested under
monotonic loading conditions. In all
cases, failure occurred after a consid-
erable number of fully reversed cycles
and slip magnitudes larger than 5 mm
(0.2 in.).
3. The shear resistance forms an
elasto-plastic, unconfined mechanism
with irrecoverable slip. The slip could
only be recovered if the loading direc-
tion is completely reversed.
4. The shear resistance for connec-
tions with mild steel continuity rein-
forcement is mainly provided by two
components: (1) interface friction and
(2) resistance of the continuity bar to
deformation. A model based on these
mechanisms was introduced and
agrees well with experimental results.
S. The shear resistance of connec-
tions with post-tensioning, using either
strands or bars, is mainly provided by
friction at the drypack grout-to-panel
interface. Using prestressing increases
the vertical stress on the connection,
thus increasing the overall frictional re-
sistance. A model was introduced and
agrees well with experimental results
at the limit states of cyclic behavior.
6. The cyclic behavior of connec-
tions with shear keys can be predicted
using models proposed previously for
monotonic loading conditions.
RECOMMENDATIONS
1. The findings of this research l •
and the companion programs·-I] have
defined the behavior and capacity of
typical and new horizontal connec-
tions for precast concrete loadbearing
shear wall panel systems. These re-
sults are a significant step towards de-
velopment and acceptance of precast
concrete wall panel systems in seismic
zones.
2. Typical connection details cur-
rently used in practice for precast wall
May-June 1996
panels exhibit stable hysteretic behav-
ior under reversed cyclic shear load-
ing. All connection configurations in-
vestigated in this program exhibited
sustained shear capacity for many cy-
cles of loading and slip magnitudes in
excess of 5 mm (0.2 in.).
3. The use of grouted shear keys
significantly increased the shear resis-
tance of the connection and limited
horizontal slip. Both of these attributes
are desirable to overall seismic perfor-
mance of this structural system.
Therefore, the use of shear keys is rec-
ommended in precast wall panel struc-
tures in seismic zones.
4. The elasto-plastic behavior of the
shear resistance mechanism provides a
source of energy dissipation. How-
ever, the slip mechanism is not recov-
erable, and accumulation of slip over
the height of the structure could lead
to structural instability. Therefore, if it
is desired to utilize horizontal slip as
an energy dissipation mechanism, the
slip must be constrained to avoid in-
stability.
S. Further research is needed to de-
fine the response of the structural sys-
tem under earthquake conditions.
Using the connection behavior defmed
in these research programs,,"I. analyti-
cal andlor experimental studies of
structural response must be performed
to ensure overall satisfactory seismic
performance.
ACKNOWLEDGMENT
This experimental program was
conducted at the University of Mani-
toba, Winnipeg, Canada, with finan-
cial assistance from the Natural Sci-
ences and Engineering Research
Council of Canada (NSERC), Con-
Force Structures Ltd., Winnipeg, Man-
itoba, and LaFarge Construction Mate-
rials (formerly, Supercrete Ltd.),
Winnipeg, Manitoba. The help of
graduate student Nolan Domenico and
laboratory technicians Ed Lemke and
Moray McVey is greatly appreciated.
APPENDIX - NOTATION
Ac = total cross-sectional area of con-
nection
As =total cross-sectional area of con-
tinuity bars crossing connection
Ie'= cylinder compressive strength
of panel concrete
I; =equivalent standard cylinder
compressive strength of drypack
grout
1;= cube compressive strength of
drypack grout
=
Is tensile stress increase in mild
steel continuity bars
=
J, nominal yield strength of mild
steel continuity bars
Fb = compressive force in mild steel
continuity bars due to applied
gravity load
=
Vb frictional resistance provided by
clamping action of bar
Vj1 :::: shear resistance component pro-
vided by friction due to net
gravity load acting on drypack
"l2 : : shear resistance component pro-
vided by friction due to clamp-
ing action (kinking mechanism)
V"l :::: shear resistance component pro-
vided by flexural mechanism in
continuity bars
VII2 =direct shear resistance compo-
nent provided by kinking mech-
anism
=
V" frictional resistance provided by
load normal to connection
=
Vp frictional resistance provided by
post-tensioning
=
Vr shear resistance of connection
(for design)
V,. =ultimate shear resistance of con-
nection after failure of drypack
grout
J.I. =coefficient of friction
0",. =compressive stress on connec-
tion due to normal loads
0",,1 = net compressive stress normal to
connection due to gravity loads
0"112 = compressive stress normal to
connection due to clamping ac-
tion
O"p = compressive stress normal to
connection due to post-tension-
ing
REFERENCES
1. ACI Committee 318, "Building Code
Requirements for Reinforced Concrete
(ACI 318-89)," American Concrete In-
stitute, Farmington, MI, 1995.
2. Canadian Standards Association
(CSA), "Design of Concrete Structures
for Buildings (CAN-A23.3-M84),"
Rexdale, Ontario, Canada, 1984.
3. International Conference of Building
B
 Officials, Uniform Building Code,
Pasadena, CA. 1994.
4. SEAOC, "Recommended Lateral
Force Requirements and Commen-
tary," Seismology Committee of the
Structural Engineers Association of
California, Sacramento, CA. 1990.
5. Federal Emergency Management
Agency, NEHRP Recomme1llkd Pro-
visions for the Development of Seis-
mic Regulations for New Buildings,
1991 Edition, National Earthquake
Hazard Reduction Program, Washing-
ton, D.C.
6. PCI Design Handbook, 4th Edition,
Precast/Prestressed Concrete Institute,
Chicago, IL. 1992.
7. Connections for Precast and Pre-
stressed Concrete, 2nd Edition, Pre-
cast/Prestressed Concrete Institute,
Chicago, II., 1977.
8. Soudki, K. A., ''Behavior of Horizon-
tal Connections for Precast Concrete
Loadbearing Shear Wall Panels Sub-
jected to Large Reversed Cyclic De-
formations," Ph.D. Thesis, University
of Manitoba, Winnipeg, Manitoba,
April 1994, 670pp.
9. Soudki, K. A., Rizkalla, S. H., and
LeBlanc, B., "Horizontal Connections
for Precast Concrete Shear Walls Sub-
jected to Cyclic Deformations - Part
1: Mild Steel Connections," PCI
JOURNAL, V. 40, No.4, July-August
1995, pp. 78-96.
10. Soudki, K. A., Rizkalla, S. H., and Di-
akiw, R. W., "Horizontal Connections
for Precast Concrete Shear Walls Sub-
jected to Cyclic Deformations - Part
2: Prestressed Connections," PCI
JOURNAL, V. 40, No.5, September-
October 1995, pp. 82-96.
11. Foerster, H. R., Rizkalla, S. H., and
Huevel, 1. S., "Behavior and Design of
Shear Connections for Loadbearing
Wall Panels," PCI JOURNAL, V. 34,
No. I, January-February 1989, pp.
102-119.
12. Hutchinson, R. L., Rizkalla, S. H.,
Lau, M., and Huevel, J. S., "Horizon-
tal Post-Tensioned Connections for
Precast Concrete Loadbearing Shear
Wall Panels," PCI JOURNAL, V. 36,
No.6, November-December 1991, pp.
64-76.
13. Serrette, R. L., Rizkalla, S. H., At-
tiogbe, E. K., and Huevel, 1. S., "Mul-
tiple Shear Key Connections for Pre-
cast Shear Wall Panels," PCI
JOURNAL, V. 34, No.2, March-April
1989, pp. 104-120.
14. West, J. S., "Behavior of Horizontal
Connections for Precast Concrete
Loadbearing Shear Wall Panels Sub-
9 PCI JOURNAL
jected to Large Reversed Cyclic Shear Fig. 7(b). Failure mode for Connection
Loading," M.Sc. Thesis, University of PTB.
Manitoba, Winnipeg, Manitoba, Au-
gust 1993, 319 pp. Fig. 7(c). Connection SK immediately
after cracking.
Fig. 1(a). Connection DP - Drypack Fig. 7(d). Connection SK at failure.
grout only.
Fig. 8. Stages of cyclic shear-slip
Fig. 1(b). Connection SK - Shear connection behavior.
keys.
Fig. 9. Effect of cyclic load.
Fig. 1(c). Connection RW -
Reinforcing bar with welded Fig. 10. Effect of mild steel continuity
connection. reinforcement.
Fig. 1(d). Connections PTS and PTB Fig. 11. Effect of post-tensioning.
- Post-tensioned using strands or
bars. Fig. 12. Effect of shear keys.
Fig. 2. Test specimen. Fig. 13. Initial frictional resistance
(a) Schematic provided by net gravity load acting on
(b) Photographic Connection RW.
Fig. 3. Test setup. Fig. 14(a). Resistance of continuity bar
(a) Schematic to shear deformations: flexural
(b) Photographic mechanism.
Fig. 4(a). Applied reversed cyclic Fig. 14(b). Resistance of continuity bar
loading history - Load control. to shear deformations: kinking
mechanism.
Fig. 4(b). Applied reversed cyclic
loading history - Slip control. Fig. 15. Comparison of predicted vs.
measured response for connection
Fig. 5(a). Hysteresis loops of shear with mild steel continuity bars.
resistance vs. slip behavior for
Connection DP. Fig. 16. Comparison of predicted vs.
measured response for connection
Fig. 5(b). Hysteresis loops of shear with post-tensioned strands
resistance vs. slip behavior for
Connection SK.
Fig. 5(c). Hysteresis loops of shear
resistance vs. slip behavior for
Connection RW.
Fig. 5(d). Hysteresis loops of shear
resistance vs. slip behavior for
Connection PTS.
Fig. 5(e). Hysteresis loops at shear
resistance vs. slip behavior for
Connection PTB.
Fig. 5(t). Shear resistance vs. slip
static behavior for Connection PTB-S.
Fig. 6. Correlation between loss of
shear resistance and crushing of
drypack grout.
Fig. 7(a). Failure mode for Connection
RW.
Table 1. Outline of experimental program.
Type Connection description Test method
DP Plain surface, drypack grout only • Cyclic
SK Shear keys only • Cyclic
RW Reinforcing bar welded to steel angle • Cyclic
PTS Post -lenSiooed strand • Cyclic
PTB Post-tenSiooed bar Static Cyclic
Total nmnber of shear specimens = 6 1 5
• Static ItStI pcrfonned in previous studies at the University of Manitoba (Refs. 11-13).
 Upper Panel

dfYpack

Lower Panel

Fig. lea). Connection DP - drypack grout only.

Upper Panel

drypacked
shear keys (5)

Lower Panel

Fig. 1(b). Connection SK - shear keys.

 Upper Panel

reinforcement
~ (25M, 400W)
drypack

weld
steel
angle

Lower Panel

Fig. l(c). Connection RW - rebar with welded connection.

 Upper Panel

Lower Panel

strand
or bar

.,

j .."" ., ~ '·L·~

Fig. led). Connections PTS and PTB - post-tensioned using strands or bars.
 -----------3800 mm----------~
...~ Simulated

I Gravity Load

Cyclic
1000 mm
:
I
I
Top Panel
Region
l
1400 mm

~I~:~J
Shear Force

I I
1100 mm I Bottom Panel I

L
I I
I I
Panel Width = 150 mm ___9_0_0__.....I~
.L.1I---
Continuity
Reinforcement
1...~--12 00 mm .1

(a) Schematic

(b) Photographic

Fig. 2. Test Specimen.

 R eoct;on Wall
,-,--
I
I
I Gravity Load \
Simulator
I Lateral Broce
,----

I
I
I
rh rh
I \ ~ / I
I
I L
l--~lbL II
,
I
.-- I----
I
I 0
0
I
I
I
t1 Loading Yoke
I
I
I
/;
I
I
I
I
I
I
I
/
o
I
f----J
r H\
Lo
/pqri J~
I
~~ ~

~
I f---
I
Cross Beam) Actuator
I
f---
I
(/ "--- f--- '---1 ~ ~I Lateral Brace
I

Structural Floor
\\ Reaction Abutment

(a) Schematic

(b) Photographic

Fig. 3. Test Setup.

 800~----------------------------------------,
push direction

600 .................................................................................... .
Up to
400 ................................................. . ......... , initiation ...
of slip

z=- 200 .............. A.... ~····~··· ... ... .. .. ...... .

"t:J
CIS
o
..J o V\ A A
"t:J VVV
.!
8: -200 ................ ~ ... ~ ...y.... .. ... ... ... .. ..
<
-400 ........................................................................... .

-600 .................................................................................... .
pull direction

-800~----------------------------------------~
Increasing Cycles •

Fig. 4(a). Applied reversed cyclic loading history - load control.

8~------------------------------------------,
push direction

6 ................................................................................... .

4 ........................................................ .•. . , .. Up To
Failure ••

-E .............j ..............................
-
E

:
2

a ~ AAJ
---t---
"t:J
.! VVV
:i -2 ..................Y... ... .. .. .. .. .. .. . .... " ............. .
<
-4 ................................................... , ..... " .............. .

-6 ..................................................................... '" ............ .

pull direction

-8~--------------------------------------------~
Increasing Cycles •

Fig. 4(b). Applied reversed cyclic loading history - slip control.

 (inches)
-0.6 -0.4 -0.2 o 0.2 0.4 0.6
1000 +-------~------_r------_r------_+------~------__r225

800 •..•..•..••.••••..........••• , ...•.......... , .•........• • I

-... ~
•
.... -.-.-.- .... I
~.- ........ -.- 180

- ... · ... ... ..

• • I • •
, • t • I

Z 600 · ...........................
.. _.- -......... ,-- _ .............................................................................. -_ . . 135

-
• • I • •

..
I • , • •

~ • I I
. , I •

400 .............. i·············· i·············· .: ...............:...............:.............. 90

Q)
(,) .· ..
•

•
I

•
.
•

I
.•

•
..
•

200 .............. ;..... . ... . ............. -- 45

s::::
....ca
.!! 0 o
-
en
Q,

tn
Q)
It: -200 L..:.kl=::l:~W. _ _"'"~~·:.r· ......'-............. -45 -
32

...
ca -400 .............. : .............. ~ ... -.......... ~ .. _. --........ ~ .......... --.. -:.......... --. . -90
Q)
.c
·: .: .
: Connection OP:
t/) -600 · .
.............. ; .............. ; .............. ; .............. Orypack only·········· -135

-800 -180
··
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . '"I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1- . . . . . . . . • • . . . . . . . . . . . . ..

I
..
I

•
..
•

•
.
I

.
I
.
•

.
t

-1000 +-------~-------+------_+------~------~r-------+~25

-15 -10 -5 0 5 10 15
Applied Slip Magnitude (mm)

Fig. 5(a). Hysteresis loops of shear resistance versus slip behavior for Connection DP.
 (inches)
-0.6 -0.4 -0.2 o 0.2 0.4 0.6
1000 +-------~-------4--------~------+_------_r------__+225

.. .. .. .. .. .. .. .. . . .. . .... . . ........................... ~.. . . . .. . . .. . .. . . .... . ........................ ..:- ...........................:............................ 180

800 · ... .
···
-
Z 600 ............................ . ...................................................... . ........................ -.-· .......................... -.-"
, .
··,,
. ...................... .
.. 135

-
.:.::::
Q)
0
400 .........:................:. ........ .....
· .
90

c
...,ca
.!
200

0
.................. ---------------. 45

o
-
I I)
Q.
tn
Q)
n=... -200 -45
-
:i

ca -400 . .......................... ~- ............................ :. . . . . . . . . . . . . . -90

Q)
.c · .
en -600 .......................................... _...
··· ... .. -.---....... 1': Shear Connection SK:
Keys -135
· .
-800 .......................................................................... .
···
·
...
.
····
. . ............................................................ -,-. ........................ . .
... -180

-1000 +--------+-------4--------~------4-------_+--------+-225
· .
-15 -10 -5 0 5 10 15
Applied Slip Magnitude (rnm)

Fig. 5(b). Hysteresis loops of shear resistance versus slip behavior for Connection SK
 (inches)
-0.6 -0.4 -0.2 o 0.2 0.4 0.6
1000 +-------~------_r------_+------_+------~------__t225

800 · . . . .
............................ 9 ............................ " ....................................................................................... - .......................... -
180
·
- . . . .
• • • I I
• • I • •

600 ............................ ; .............................

•

~
•

............................ ~ ........................... ..: .............................:......................... ..

I I •

135
Z
-
I I I • •
I I • • •

~ I •
,
•
I
•
I •
•

400 .............. ; .............. -: .............. -:. . . . . . . . . . . . . . ....:. . . . . . . . . . . . . . 90

CI)
0
: : ~~ .
c 200 ....... -.. ----:--.--.-.~
.. - ~.~~ 45
...,C'G
.~ 0
- o
I /)
Q.
fn
CI)
0:: -200 :O:1FEH~tIi~tililijJi.~· -.... -.:............ -- - -45
:i:

...
C'G -400 . -.... -. -. -. -~. -... -.... -. - .~ .... -. -. -.... -.~ ... --- -. -- .. - -90
CI)
.c · Connection RW:
tn -600 . -..... - .. -. -. : -..... -. --.... : -... -...... -. -:" ... -.... -. -. Welded Rebar -.. -... -.. -135
·
•
. •
. I I •

-800 ··
I

. .. I

.. . .

..
............................ ., ............................................................................................................................................. . -180

-1000
·•

+-------~-------4--------~------4-------_+------__+-225
.. •
. •
. I
. •

-15 -10 -5 o 5 10 15
Applied Slip Magnitude (mm)

Fig. 5(c). Hysteresis loops of shear resistance versus slip behavior for Connection RW.
 (inches)
-0.6 -0.4 -0.2 o 0.2 0.4 0.6
1000 +--------+------~--------r_------;_------_+------__+225

· . . . .
............................ ,. ................................................................................................................................................... .. 180
800
·· .. . .. ..
-
• .. I • ,

..
............................ , ............................ , ......................................................... ................................. -.- ......................... .
• , • I •

600 135
I • • • •

Z
.
.. ..
- · .
, • f l '

.llI:: .·~~~~t·~~~~~ . .... . ........................... 90

400
CI)
(.)
c 200 .............. 1·······~·····dJ
.. ~~~~Tl 45
....tnCO
.- 0 o
-
I /)
Q..

tn
CI)
c::: -200 -45
-
:i

"-
CO
CI)
.c
-400 .. ~
··············:···········c .. W~~~i ..........·.... ........·....
~ -90

tJ) -600 .................................................

·
I

. " '

. . .
............................................................................................................................
•

-135
·:: :
.
Connection PTS:
.
-800 .............. : .............. :' ........... ":'" ... Post-tensioned Strands ........ . -180
·· .. ..
-1000 +--------+------~--------+-------4--------+------__+-225

-15 -10 -5 o 5 10 15
Applied Slip Magnitude (mm)

Fig. 5(d). Hysteresis loops of shear resistance versus slip behavior for Connection PTS.
 (inches)
-0.6 -0.4 -0.2 o 0.2 0.4 0.6
1000 +-------~------_r------_+------_+------~------__t225

. . ..
-_ .................................................................................
800
···· . . 180
............... .
. .
-
Z 600
.
.
..
··
.
.................................. ,·...............................................................
.
..
. 135

-
~

GJ
(.)
400 ............................ . .......................... ........... .
~

· 90

200
c
....ca
.~ 0
45

0
-
I /)
Q.
:i:
tn
GJ
0:: -200 -45
...
ca -400
GJ ..
. ···
.... _- ... -_ ...... _--_ ....... - ......... .. -90
.c
en -600 ·. · .
.
................................................................................................
· -135
: : Connection PTB:
-800 . - - --- . - .. - . - . : -- - --. - . -- - .. - ~ -- .. _.. - . -- -- - . - . - .... Post-tensioned Bars .... - . - . - . - -180
-1000
··· ...
-225
-15 -10 -5 0 5 10 15
Applied Slip Magnitude (mm)

Fig. See). Hysteresis loops of shear resistance versus slip behavior for Connection PTB.
 (inches)
-0.20 0.20 0.59 0.98 1.38 1.77 2.17 2.56 2.95
1000 ~~--+-----~----~----~----~----~-----+-----T225

900 .. . ...... . ...... ~ .....................:. .................... ; .................... ~ .....................:. .................... ; .................... ~...... . . ...... . .. 202

-
I •
I •
• •
I I
• • • • I
, , I • t
• • • • I
, • I I I

800 .. . .. .... . ...... ~ ......................:...................... ~ ..................... ' ..................... -:- .................... ; .................... ~ .. . .. .. . .. . .. .. .. 180
Z ·· .. .. ..
-
• I • I

~ • • I •

700 . . . .... . ...... .................... -:. . . . . . . . . . . ................... , ...................... :- .................... ; . . . . . . . .... .; . . . . . .. . .. . .

~ 157
• , I • I • •
CI) • I I • • • •

(,)
c 600
•

.................. -:-................
•
..................... : ..................... -:- ...................:- .................... ; ...................................... .
I
I
•
•
•
I
•
,
..
•
•

135
...,co I

· ..
•
.. •
.. •
.. •

..
•

. -.9-
I I)
.. ...... .......................:- .................... .................... .....................:...................... .................... ................ ..
I • • • • • •

.! 500 . .. ......

-
~ ~ { ~ ~
112
U)
CI)
,

,
·· ,

,
.. ..•

I
.. ,

I
.. ,

•
.
.
•

I
.. I

•
oX

400 . . . .. . ... ...........:........... : .......... ...........:........... ! .......... . ........

~ ~ ~ 90
0::: ·: .: .
:
.: .: .: .
: '---
a...
CO 300 .. . . .... . ...... ~ ......................~ .................... ; .............. --- ~ .... -...... -.... --:- ...... -............ ; ............ -.. -.. ~ . . . . . . . .. . . 67
CI) · . . . . . .
.c : : : : Connection PTB-5: :
U) 200 •.......• ; .....•.....:- .......... ; .......... ;. Post-tensioned Bars .......•.......... 45
·
•
. . • • • I

100 .. ... . ... : ........... .......... : .......... ........... .......... : .......... :. . . . . . . . . .

~ ~ ~ 22
··•
.. .
. •
.
.
.
. •
.. ..• • I •

0 ~~--+-----4-----~----~----~----~-----+-----+0

-5 5 15 25 35 45 55 65 75
Applied Slip Magnitude (mm)

Fig. 5(t). Shear resistance versus slip static behavior for Connection PTB-S.
 (inches)
-0.08 0.00 0.08 0.16 0.24 0.31 0.39 0.47
1000 225

800 ......................
·· ..
. .
. .
.
. .
.. .................................................................................................................. .or" .................. .. 180

-
I I I • •
• • I • •
I I , • •

600 .. " .... .. .. .. ...... .. .. .................... ,• ......................... •r ........ .............. -.-"

•
.................... ,I ...................... •
-,r .................... .. 135
Z
· . . . .
-
I • • I I
• • I I •

~
. , ........................ , ....................... -.- ....................... , .............................................. .
I I • • •

400 . . . . .. . . . . . . • I • I I
90
Cl)
(J
.•
,

..
•
I

..
I
I

..
I
I

s::: 200 45
....ca Ii)
a.
.-
tn 0 0
tn
Cl)
0:: -200 -45
-
:.iii:

'-
ca -400
, .. ,
. _ .. _4 ____ .. _ .. _ _ _ _ .. "' _ _ _ _ .. _ _ _ _ _ _ .. ' .. _ _ _ _ _ .. __ .... _.1 .. _ _ _ _ _ _ _ _ _ - - ' - .. - - - - - - - - - -
-90
Cl) ·
• I
. . , • I •

.c
en -600 ........................
•
I

~
. . I
•
._.................. _.. _ .................. .
. -................................................
.. .
I
I
I
•

. -135
, • • • I

• • I • •

. , ,
. ,
• • • I

-800 ....................................................................
I •
-........ _-
• I •
-180
· . . . .
• • I • •
• • I I I

-1000 ~----~----~.~----~.------~.------~.------~.----~ -225

-2 o 2 4 6 8 10 12
Reduction of Drypack Thickness (mm)

Fig. 6. Correlation between loss of shear resistance and crushing of drypack grout.
Fig. 7(a) Failure mode for Connection RW.

Fig. 7(b). Failure mode for Connection PTB

Fig. 7( c). Connection SK immediately after cracking.

Fig. 7(d) Connection SK at failure.

 IStage "I /Stage ,,"

--
IStage II .......... -
"
-- Stage I - Elastic behavior before
initiation of slip.
...... Stage II - Inelastic behavior before
significant deterioration of drypack.
Stage III - Inelastic behavior after
significant deterioration of drypack

Applied Slip Magnitude

Fig. 8. Stages of cyclic shear-slip connection behavior.

 (inches)
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
1000 225

900 I PTB
-PTB-S
I PTB-S
202

- 800 180
Z (failure at 70 mm slip)
~
- 700 157
Cl)

..
CJ
C
C'G
.-
tn 500
600 135

112
-.9-
CI)

tn
Cl)
0:: 400 90
-
..:.::

...C'G 300
PTS
(envelope of cyclic response) 67
Cl)
.c
en 200 45

100 22

0 0
-5 0 5 10 15 20 25 30
Applied Slip Magnitude (mm)

Fig. 9. Effect of cyclic load.

 (inches)
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
1000 225

800 180
I--DP
RWI
-
Z 600 /RW 135

-
..:.I:
Cl)
CJ
400
K:_--\_\
90

c 200 45
..,
ftS
.~ 0 0
-
I I)
Q.

tn
Cl)
~ -200
- - . . . . -----1
"op -45
-
:i:

Lo .-/
ftS -400 -90
Cl)
.c
UJ -600 -135

-800 -180

-1000 -225
-15 -10 -5 0 5 10 15
Applied Slip Magnitude (mm)

Fig. 10. Effect of mild steel continuity reinforcement.

 (inches)
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
1000 225
-PTS PTS
800 180

-
Z 600
-PTS
- -DP
I 135

-
~

Cl)
(.)
400 90

200 45
C
..,ns
.!!! 0 0
-
I/)
a.

~
In
Cl)
-200 -----
DP -45
-
:i:

~
ns -400
Cl)
-90
.c
t/) -600 -135

-800 -180

-1000 -225
-15 -10 -5 0 5 10 15
Applied Slip Magnitude (mm)

Fig. 11. Effect of post-tensioning.

 (inches)
-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
1000 225

-
Z
800

600
I
-
SKI
-DP
.~ /K 180

135

-
~
400 90
Go)
CJ
c:
ns
+J
200 ----\ --. 45
-
I I)

.! 0 0 a.
tn
Go)
0::: -200 ---...------
"bp -45
-
:.i

...ns -400
~
-90
Go)
.c
tJ) -600 -135

-800 ~ -180

-1000 -225
-15 -10 -5 o 5 10 15
Applied Slip Magnitude (mm)

Fig. 12. Effect of shear keys.

 Continuity
_____ Reinforcement
(25M,400W)
v
---..
..... ...-..-------------...--
,I lilt' 11111 11111 11111 It' t I

Fb q" Fb

J
where:

Vf1 = friction due to net gravity load

~ = coefficient of friction
cr n1 = vertical stress acting on drypack
due to net gravity load
A~ = Pg -2Fb
Ac
Pg Pg = applied gravity load
Fb = force in continuity bar due to
gravity load
Ac = cross-sectional area of
connection

Fig. 13. Initial frictional resistance provided by net gravity load acting on Connection RW.
 Continuity
. / Reinforcement
(25M,400W)

V ...
....
, Vh-J
I
V-
",M
a-
",M
,,
Vh-J

drypack
/
M
Vf-Vh s-:J
M~
L

V
-
...
~

where:
Vh1 = shear resistance due to
flexural mechanism in bar
M = end moment in bar, limited
to plastic moment
L =
deformed length

Fig. 14(a). Resistance of continuity bar to shear defonnations: flexural mechanism.

 Contill.lity
- - - -Reinforcement
(25M,400W)

V.., vf2 ~
.....
a~~A~: IIIJ~~~~j~F
I

• aF
b
by aF
b
by ,
t (Jra t
t'H f
"itifffttl.""itlt!il ..
drypack - - . . . ---IIIiIifZ....... - - -
tF. IIla~ I'H It. f I ( V12 = Jl O'n2 Ac )
aF aFb aF aFb

"II",
,
bY~aF
If.! bx (Jra

II ,II,VII "II" ,
I2
""-+
bY~aF
~ bx
( V h2 = 2 [t.Fbx 1)
where:

Vf2 = friction due to clamping action

Vh2 = direct shear resistance due to kinking
J.! = coefficient of friction
O"n2 = stress on drypack due to clamping
= 2 [L\Fby ]
Ac
L\F b =
increase in bar force due to kinking
L\F by = vertical component of L\Fb
L\F bx = horizontal component of L\Fb
Ac =
cross-sectional area of connection

Fig. 14(b). Resistance of continuity bar to shear deformations: kinking mechanism.

 (inches)
-0.08 0.00 0.08 0.16 0.24 0.31 0.39 0.47
600 +------+------+------+------r-----~------~----_t135

_ 500 112
Z IVr =416 kN I
-
~

CI) 400
(93.5 kips)

90
(.)
c
S
\measured -
en
.!! 300 67 Co
In
CI)
0::
-
:i:

... 200
tCS
CI)
Ivu =216 kN (48.6 kips) I 45

.c
U)
100 22

o +-----~~----~------~-----+------+_----~------_+o

-2 o 246 8 10 12
Applied Slip Magnitude (mm)

Fig. 15. Comparison of predicted versus measured response for connection with welded mild
steel continuity bars.
 (inches)
-0.08 0.00 0.08 0.16 0.24 0.31 0.39 0.47
600 135
measured
model

_
Z
500 '" IVr =460 kN (103.5 kips) I
- . . . -- - - - - - - -- ---
112

-
..lII:

(1)
(.)
400 90

~
C
CO
.~ 300 67
-
II)
a.
U'J
(1)
ex:
-
~

'-
CO
200 IVu =216 kN (48.6 kips) I 45
(1)
.c
U)
100 22

0 0
-2 0 2 4 6 8 10 12
Applied Slip Magnitude (mm)

Fig. 16. Comparison of predicted versus measured response for connection post-tensioned with
strands.